No-assisted N2O decomposition over Fe-based catalysts: Effects of gas-phase composition and catalyst constitution

Article abstract of DOI:10.1006/jcat.2002.3559

The decomposition of N₂O is strongly promoted by NO over Fe catalysts supported on zeolites (ex-framework FeMFI catalysts, sublimed Fe/ZSM-5, ion-exchanged Fe-ZSM-5, Fe-beta, and Fe- USY) and conventional supports (Fe/Al₂O₃ and Fe/SiO₂) in a wide temperature range (550-900 K). Mixtures of NO and O₂ in the N₂O- containing feed lead to the same promoting effect as NO only. The promoting effect of NO is catalytic, and in addition to NO₂, O₂ is formed much more extensively at lower temperatures than in the absence of NO. The promotion effect only requires low NO concentrations in the feed, with no significant improvements at molar NO/N₂O feed ratios higher than 0.25. No inhibition by NO was identified even at a molar NO/N₂O feed ratio of 10, suggesting different sites for NO adsorption and oxygen deposition by N₂O. The latter sites seem to be remote from each other. Multitrack experiments on ex-framework FeZSM-5 show that release of oxygen from the catalyst surface during direct N₂O decomposition is a rate-determining step, due to the slow oxygen recombination, which is favored by high reaction temperatures. NO addition promotes this oxygen desorption. Adsorbed NO accommodates oxygen from N₂O and the formed adsorbed NO₂ can react with a second oxygen from the neighboring site, thus accelerating the recombination of oxygen from N₂O. Less than 0.9% of the Fe seems to participate in this promotion. Adsorbed NO may even facilitate the migration of atomic oxygen to enhance their recombination. A model is proposed to explain the phenomena observed in the NO-assisted N₂O decomposition, including NO₂ decomposition.

Full text of DOI:10.1006/jcat.2002.3559

Journal of Catalysis 208, 211–223 (2002)
doi:10.1006/jcat.2002.3559, available online at http://www.idealibrary.com on
NO-Assisted N2O Decomposition over Fe-Based Catalysts: Effects
of Gas-Phase Composition and Catalyst Constitution
Javier Pe´rez-Ram´ırez,1 Freek Kapteijn, Guido Mul, and Jacob A. Moulijn
Industrial Catalysis, DelftChemTech, Delft University of Technology, Julianalaan 136, 2628 BL, Delft, The Netherlands
Received November 21, 2001; revised February 7, 2002; accepted February 7, 2002
FeZSM-5 catalyst prepared via an ex-framework method in
the N2O decomposition reaction (16). The ex-framework
catalyst (on a per Fe basis) shows a significantly higher ac-
tivity than catalysts prepared via other procedures, such
as liquid (aqueous)- or solid-ion exchange, or sublimation.
Also the stability of this particular catalyst in simulated
tail-gas mixtures from nitric acid plants and fluidized-bed
combustors is excellent (16, 17).
An intriguing general feature of FeZSM-5 catalysts in
N2O decomposition is that NO significantly enhances the
activity, while the opposite effect is usually observed for
other catalytic systems, e.g., those based on noble metals
(Ru or Rh) (18, 19). This peculiar behavior of FeZSM-5
makes it very attractive to use in applications where both
N2O and NO are present, such as in tail gas of nitric acid
plants. The positive effect of NO and also of SO2 on the N2O
conversion was first reported by some of us in 1996, using a
FeZSM-5 prepared by liquid (aqueous)-ion exchange with
Fe(II) sulfate (20). At that time it was proposed that NO
in the gas phase scavenged adsorbed oxygen (deposited by
N2O during the oxidation of active sites, Eq. [1]), leading
to the formation of NO2 and regeneration of the active site
(Eq. [2]).
The decomposition of N2O is strongly promoted by NO over
Fe catalysts supported on zeolites (ex-framework FeMFI catalysts,
sublimed Fe/ZSM-5, ion-exchanged Fe–ZSM-5, Fe-beta, and Fe–
USY) and conventional supports (Fe/Al2O3 and Fe/SiO2) in a wide
temperature range (550–900 K). Mixtures of NO and O2 in the N2O-
containing feed lead to the same promoting effect as NO only. The
promoting effect of NO is catalytic, and in addition to NO2, O2 is
formed much more extensively at lower temperatures than in the
absence of NO. The promotion effect only requires low NO con-
centrations in the feed, with no significant improvements at molar
NO/N2O feed ratios higher than 0.25. No inhibition by NO was iden-
tified even at a molar NO/N2O feed ratio of 10, suggesting different
sites for NO adsorption and oxygen deposition by N2O. The latter
sites seem to be remote from each other. Multitrack experiments on
ex-framework FeZSM-5 show that release of oxygen from the cata-
lyst surface during direct N2O decomposition is a rate-determining
step, due to the slow oxygen recombination, which is favored by high
reaction temperatures. NO addition promotes this oxygen desorp-
tion. AdsorbedNOaccommodatesoxygenfromN2Oandtheformed
adsorbed NO2 can react with a second oxygen from the neighbor-
ing site, thus accelerating the recombination of oxygen from N2O.
Less than 0.9% of the Fe seems to participate in this promotion.
Adsorbed NO may even facilitate the migration of atomic oxygen
to enhance their recombination. A model is proposed to explain
the phenomena observed in the NO-assisted N2O decomposition,
N2O + ∗ → N2 + O∗
NO + O∗ → NO2 + ∗
[1]
[2]
c
including NO2 decomposition. ꢀ 2002 Elsevier Science (USA)
Key Words: N2O decomposition; NO; Fe catalysts; FeZSM-5; ex-
framework; Fe-beta; Fe–USY; Fe/Al2O3; promotion; O2 desorption;
mechanism; TAP reactor; transient kinetics.
Reduction of an oxidized site by NO as represented in
Eq. [2] should lead to a stoichiometric process in which NO
and N2O react to form N2 and NO2:
INTRODUCTION
N2O + NO → N2 + NO2.
[3]
Fe-based zeolite catalysts are currently extensively stud-
ied, because of high activity in (i) selective catalytic re-
duction of NO and N2O with hydrocarbons or ammonia
(1–8), (ii) N2O-mediated selective oxidation of benzene to
phenol (9–11), and (iii) direct catalytic N2O decomposition
(12–14), and selective oxidation of NH3 to N2 (15). We re-
cently reported the extraordinary performance of a specific
Recently we reported transient experiments using FT-
IR/MS (Fourier transform infrared coupled to mass spec-
trometry) and Multitrack (multiple time resolved analysis
of catalytic kinetics), an advanced TAP-reactor, to fur-
ther analyze the NO-assisted N2O decomposition over ex-
framework FeZSM-5 (21). It was shown that besides NO2
formation, enhanced O2 formation is involved. It was pro-
posed that the latter is either an indirect effect (electronic
or steric) of NO adsorbed on sites neighboring the N2O
1 To whom correspondence should be addressed. Current address:
Norsk Hydro, Research Centre, P.O. Box 2560, N-3907, Porsgrunn,
Norway. Fax: 47 35 923043. E-mail: javier.perez.ramirez@hydro.com.
211
0021-9517/02 $35.00
c
ꢀ 2002 Elsevier Science (USA)
All rights reserved.

´ ´
PEREZ-RAMIREZ ET AL.
212
decomposition sites, or a direct effect involving reaction ratio of 0.2 (denoted ex-[Fe,Al]MFI(b)). Following this
of adsorbed NO2 groups with neighboring oxidized sites route, ferrigallosilicate (denoted ex-[Fe,Ga]MFI) and fer-
yielding O2.
risilicate (denoted ex-[Fe]MFI) were also prepared. Addi-
In this paper it is shown that NO-assisted N2O decom- tion of NaOH to the synthesis gel induces a lower template
position is not a special feature of FeZSM-5 catalysts, but concentration and thus a larger crystal size for the final
a general phenomenon occurring over many Fe-containing catalysts (0.4 µm in ex-[Fe,Al]MFI(a) vs 2.0–2.5 µm in ex-
zeolitic and nonzeolitic catalysts. Furthermore, additional [Fe,Al]MFI(b), ex-[Fe,Ga]MFI, and ex-[Fe]MFI catalysts).
activity data on the effect of the molar NO/N2O ratio and More details on the preparation methods and characteriza-
reaction temperature on the performance of these catalysts tion of these materials can be found elsewhere (11).
is presented. Based on flow and pulse experiments, steady-
Other Fe catalysts. Fe/ZSM-5 was prepared by subli-
state and transient phenomena regarding the chemistry of
mation of FeCl3 on HZSM-5 (Degussa, Si/Al = 14.0), ac-
N2O, NO, and NO2 over Fe catalysts are investigated, and
cording to the method described by Chen and Sachtler
possible catalytic cycles in NO-assisted N2O decomposition
(2). Fe–ZSM-5, Fe-beta, and Fe–USY were prepared by
are proposed.
liquid (aqueous)-ion exchanged with diluted solutions
(0.30 mM) of Fe(NO3)3 · 9H2O (Fe–ZSM-5 and Fe–USY)
EXPERIMENTAL
and FeSO4 · 7H2O (Fe-beta). The parent zeolites were pro-
vided by Zeolyst (NH4–ZSM-5, CBV8020, Si/Al = 37.5;
Catalysts
beta, CP814E, Si/Al = 11; H–USY, CBV720, Si/Al = 15.5).
The chemical compositions of the different catalysts used
in this study are summarized in Table 1.
The exchange was carried out under vigorous stirring over
the course of 15 h. The zeolites were then filtered, washed
thoroughly, dried, and finally calcined in static air at 825 K
for 5 h. Fe/Al2O3 and Fe/SiO2 were prepared by incipient
wetness impregnation, with appropriate aqueous solutions
of iron(II) gluconate, Fe[OHCH2(CHOH)4CO2]2 · 2H2O.
Al2O3 (CK300) and SiO2 were purchased from Ketjen and
Grace, respectively. After the impregnation the samples
were dried at 363 K, followed by calcination at 493 K for
1 h and at 773 K for 5 h.
Ex-framework Fe catalysts. Isomorphously substituted
FeMFI was synthesized hydrothermally using tetrapropy-
lammonium hydroxide as the template (11). The molar
ratios between the components was H2O/Si = 45, TPAOH/
Si = 0.3, Si/Al = 36, and Si/Fe = 152. The as-synthesized
sample, in which Fe(III) is isomorphously substituted in
the zeolite framework, was calcined in air at 823 K for
10 h and was then converted into the H form by three
consecutive exchanges with an ammonium nitrate solution
(0.1 M) overnight and subsequent calcination at 823 K for
5 h. Finally, the catalyst was treated in flowing steam at
ambient pressure (water partial pressure of 300 mbar and
30 ml · min−1 of N2 flow) at 873 K over the course of 5 h,
yielding a sample denoted ex-[Fe,Al]MFI(a).
Activity Tests
Activity measurements were carried out in a six-flow re-
actor system (22), using 50 mg of catalyst (125–200 µm)
and a space velocity (GHSV) of 60,000 h−1 at atmospheric
pressure. The feed conditions used were 1.5 mbar of N2O,
0–15 mbar of NO, 0–0.2 mbar of NO2, and 0–20 mbar of
O2, with He as balance gas. Before reaction, the catalysts
were pretreated in N2O (1.5 mbar of N2O in He) at 723 K
for 1 h and cooled in that gas flow to the initial reaction
temperature. In the temperature range of 550–925 K, deac-
tivation of the catalysts was absent. N2O, N2, and O2 were
analyzed with a GC (Chrompack CP 9001) equipped with
a thermal conductivity detector, using a Poraplot Q col-
umn (for N2O separation) and a Molsieve 5A column (for
N2 and O2 separation). NO, NO2, and NOx concentrations
were determined with a chemiluminescence NOx analyzer
(Ecophysics CLD 700 EL).
In a modified synthesis route, ex-[Fe,Al]MFI was pre-
pared using the same molar metal ratios but a lower mo-
lar TPAOH/Si ratio (0.1 instead of 0.3). This was done
by adding NaOH to the synthesis gel in a molar NaOH/Si
TABLE 1
Chemical Composition of the Catalysts Used in this Study
Catalyst
Method
Si/Al or Si/Gaa
Fea (wt%)
ex-[Fe,Al]MFI(a)
ex-[Fe,Al]MFI(b)
ex-[Fe,Ga]MFI
ex-[Fe]MFI
Fe/ZSM-5
Fe–ZSM-5
Fe-beta
Fe–USY
Ex-framework
Ex-framework
Ex-framework
Ex-framework
Sublimation
Ion exchange
Ion exchange
Ion exchange
Incipient wetness
Incipient wetness
32.6
32.6
32.6
—
14.0
37.5
11.0
15.5
—
0.66
0.64
0.58
0.68
5.0
1.44
1.50
1.28
2.0
Multitrack Experiments
Experimental setup. A detailed description of the Mul-
titrack system, a TAP-reactor-like high-vacuum reactor sys-
tem suited for pulse experiments, is presented elsewhere
(23). Two different gases can be dosed to the reactor
by means of high-speed pulse valves, yielding pulses of
Fe/Al2O3
Fe/SiO2
—
2.0
a Determined by ICP–OES.

NO-ASSISTED N2O DECOMPOSITION OVER Fe CATALYSTS
213
1017 molecules within 100 µs. The reactor with the catalyst
sample is located in a high-vacuum system, and during puls-
ing the peak pressure remains below 3 Pa. The catalyst bed
(100 mg; pellet size, 125–200 µm) is packed between two
layers of inert SiC particles (particle size, 230 µm), consist-
ing of 200 and 125 mg before and after the bed, respectively.
In the reactor, the shape and composition of the gas pulse
change due to processes such as diffusion, adsorption, and
reaction. At the reactor exit the reaction products are ana-
lyzed by four quadrupole mass spectrometers positioned in
line with the reactor axis. All mass spectrometers are able
to analyze one of the components (m/e unit) in the exit gas
stream with a maximum sample frequency of 1 MHz. In this
study most analyses were carried out with the mass spec-
trometer located closest to the exit of the catalytic reactor.
As the signal-to-noise ratio of this system is excellent, sin-
gle pulses are sufficient to obtain good peak signals. This
is an important aspect, as transient phenomena may re-
main unobserved when several pulse responses have to be
averaged.
FIG. 1. N2O conversion vs temperature over ex-[Fe,Al]MFI(a) in dif-
ferent feed compositions: (ꢀ) 1.5 mbar of N2O in He, (ꢁ) 1.5 mbar
of N2O + 0.4 mbar NO in He, (ꢂ) 1.5 mbar of N2O + 0.4 mbar of
NO + 20 mbar of O2 in He; P = 1 bar; GHSV = 60,000 h−1
.
effect was obtained over the other ex-framework catalysts
(Fig. 2). For these catalysts, the trend in activity in the
N2O–He feed (ex-[Fe,Al]MFI(b) + ex-[Fe,Ga]MFI > ex-
[Fe]MFI) is maintained when NO is added to the feed mix-
ture. For the different Fe–zeolites, the promotion effect by
NO on the N2O conversion also leads to a significant de-
crease in the apparent activation energy (Table 2), which
has been estimated around the inflection of the activity vs
temperature curves by assuming a plug-flow model and a
first-order reaction in N2O.
Procedure. Several types of Multitrack experiments
were performed with ex-[Fe,Al]MFI(a). Before the mea-
surements, the ex-[Fe,Al]MFI(a) catalyst was evacuated in
the Multitrack reactor at 723 K for 3 h. The mechanistic
studies of direct N2O decomposition were carried out by
continuously pulsing pure N2O (∼160 nmol) at intervals of
2 s (“cycle time”) at different temperatures in the range
of 623–973 K and recording the masses m/e 28 (N2), m/e
32 (O2), and m/e 44 (N2O). Isothermal O2 desorption was
followed as a function of time after stopping N2O puls-
ing (usually after ∼100 pulses) in the temperature range of
673–798 K. The effect of NO on the N2O decomposition was
studied in dual-pulse experiments, in which pulses of N2O
(pure, ∼160 nmol) and NO (15 vol% NO in Ar, ∼24 nmol
NO) are sequentially fed to the reactor with a time interval
of 1 s between the pulses and a cycle time of 2 s. Isothermal
transient desorption effects after stopping N2O or NO puls-
ing were followed at different temperatures, ranging from
698 to 923 K. In these experiments, the masses m/e 30 (NO)
and m/e 46 (NO2) were additionally analyzed. The purity
of the gases was >99.985%.
Figure 3 shows the influence of the molar NO/N2O ratio
in the feed on the N2O conversion at reaction temperatures
ranging from 623 to 723 K. N2O conversion is dramatically
improved at low NO partial pressures (NO/N2O < 0.25).
1.0
0.8
0.6
0.4
0.2
0.0
RESULTS
Activity Data
Ex-framework catalysts. The N2O decomposition activ-
ity of ex-[Fe,Al]MFI(a) in different feed mixtures is pre-
sented in Fig. 1. In a N2O/He feed, the catalyst shows a
substantial N2O conversion only above 700 K. Addition
of NO (0.4 mbar; NO/N2O = 0.27) enhances the reaction
rate considerably. The N2O conversion curve is shifted to
temperatures about 100 K lower. The presence of oxy-
gen (20 mbar of O2) in the feed (N2O + NO + O2 in He)
hardly affects the catalyst activity. The same promotion
550
650
750
850
T / K
FIG. 2. N2O conversion vs temperature over (ꢃ, ꢂ) ex-[Fe,Al]
MFI(b), (ꢄ, ꢅ) ex-[Fe,Ga]MFI, and (ꢆ, ꢇ) ex-[Fe]MFI. Feed compo-
sitions: (ꢃ, ꢄ, ꢆ) 1.5 mbar of N2O in He and (ꢂ, ꢅ, ꢇ) 1.5 mbar of
N2O + 0.2 mbar of NO in He; P = 1 bar; GHSV = 60,000 h−1
.

´ ´
PEREZ-RAMIREZ ET AL.
214
TABLE 2
Apparent Activation Energy (Eaapp) of the Ex-Framework
Catalysts for Direct N2O Decomposition in the Absence or
Presence of NO in the Feed Gas
0.4
0.3
0.2
0.1
0.0
NO
app
app
Ea , N2Oa
Ea , N2O + NOb
Catalyst
(kJ mol−1
)
(kJ mol−1
)
x
ex-[Fe,Al]MFI(a)
ex-[Fe,Al]MFI(b)
ex-[Fe,Ga]MFI
ex-[Fe]MFI
161
137
141
155
106
90
90
p
NO2
100
a Feed: 1.5 mbar N2O in He; P = 1 bar; GHSV = 60,000 h−1
.
550
600
650
700
750
800
b Feed: 1.5 mbar N2O + 0.2 mbar NO in He; P = 1 bar; GHSV =
60,000 h−1
.
T / K
FIG. 4. Partial pressure of (ꢁ) NO, (ꢀ) NO2, and (ꢃ) NOx vs tem-
perature over ex-[Fe,Al]MFI(a). Feed composition: 1.5 mbar of N2O +
∼0.4 mbar of NO in He; P = 1 bar; GHSV = 60,000 h−1. Equilibrium com-
position of NO and NO2 (Eq. [4]) is represented by dashed lines (see text
for details).
From this value up to NO/N2O = 10, little effect on N2O
conversion is noticed.
The NO, NO2, and NOx profiles provide further infor-
mation on the effect of NO on the catalytic performance.
In a N2O + NO/He feed, the formation of NO2 over ex-
[Fe,Al]MFI(a) increases as a function of reaction tempera-
ture (Fig. 4), reaching a maximum at 650–675 K, i.e., around
the inflection point of the N2O decomposition activity curve
(Fig. 1). Above this temperature, NO2 formation decreases
before completely disappearing at 775 K. The total NOx
level is constant in the temperature range investigated, indi-
cating that NOx is not converted to N2 or N2O. The amount
of NO2 formed is beyond the thermodynamic equilibrium
of the NO and O2 reaction (Eq. [4]), suggesting that the
rate of the N2O and NO reaction is higher than the decom-
positionrateofNO2 overthecatalyst. Theequilibriumcom-
position of NO2 was calculated in two ways. If the oxygen
present at a certain temperature (generated during N2O
decomposition) is used, the short dashed line in Fig. 4 is
calculated. Equilibrium composition of NO and NO2 was
also determined assuming an O2 availability correspond-
ing to what the amount would be if N2O conversion were
100% (p(O2) = 0.75 mbar). These theoretical partial pres-
sure profiles of NO and NO2 are displayed in Fig. 4 by the
long dashed lines. The occurrence of the N2O and NO re-
action should be attributed to the catalytic performance of
ex-[Fe,Al]MFI(a), since no NO2 formation was observed
over an inert material (SiC) in the same feed composition
(not shown).
ꢀ
2NO + O2 2NO
[4]
ꢁ
2
The profiles for ex-[Fe,Al]MFI(a) in a mixture of N2O +
NO + O2/He are shown in Fig. 5a. The obtained NO/NO2
ratio at low temperatures (∼575 K) is lower than in the ab-
sence of O2. In the presence of oxygen a larger fraction of
NO in the feed mixture is oxidized to NO2 (compare pro-
files in Figs. 4 and 5a). Again, the amount of NO2 produced
by the N2O and NO reaction is beyond the thermodynamic
equilibrium in Eq. [4] (taking 0.4 mbar of NO and 20 mbar
of O2 in He as input values in the calculation), while the
presence of large amounts of oxygen shifts the theoreti-
cal equilibrium composition to higher NO2 concentrations.
However, comparing the profiles in Figs. 5a and 5b indicates
that oxidation of NO to NO2 occurs to a lesser extent in the
absence of N2O. In the NO oxidation with O2 (Fig. 5b) the
equilibrium composition of NO2 was reached above 650 K
and was not exceeded, which somehow indicates that the
behavior observed in Figs. 4 and 5a is caused by a reaction
between NO and N2O.
1.0
0.8
0.6
0.4
X
0.2
0.0
0.0
0.5
1.0
1.5
2.0
10.0
Molar NO/N2O feed ratio
In Fig. 6 the amount of NO2 produced (solid diamonds)
and O2 produced (open circles) is plotted vs the amount of
N2O reacted to N2 at 700 K for different inlet partial NO
pressures. These results clearly show that the stoichiometric
reaction of N2O with NO to NO2 and N2 (Eq. [3]) cannot
FIG. 3. N2O conversion vs the molar NO/N2O feed ratio over ex-
[Fe,Al]MFI(a) at different temperatures: (ꢈ) 623, (ꢁ) 648, (ꢅ) 673, (ꢂ)
698, and (×) 723 K. Partial N2O pressure was fixed at 1.5 mbar and partial
NO pressure varied from 0 to 15 mbar. P = 1 bar; GHSV = 60,000 h−1
.

NO-ASSISTED N2O DECOMPOSITION OVER Fe CATALYSTS
215
(a)
(b)
0.4
0.3
0.2
0.1
0.0
0.4
0.3
0.2
0.1
0.0
NO
NO2
NO
NO
x
x
p
p
NO2
NO2
750
550
600
650
700
750
800
550
600
650
700
800
T / K
T / K
FIG. 5. Partial pressure of (ꢁ) NO, (ꢀ) NO2, and (ꢃ) NOx vs temperature over ex-[Fe,Al]MFI(a). Feed composition: (a) 1.5 mbar of
N2O + 0.4 mbar of NO + 20 mbar of O2 in He, and (b) 0.4 mbar of NO + 20 mbar of O2 in He; P = 1 bar; GHSV = 60,000 h−1. Equilibrium
composition of NO and NO2 (Eq. [4]) is represented by dashed lines.
solely explain the promotion effect observed in Figs. 1 and that NO2 decomposition and NO oxidation by O2 (Eq. [4])
2. Besides inducing the formation of NO2, the presence of occur fairly fast. This result suggests that NO2 decomposi-
NO in the N2O-containing feed also significantly enhances tion might contribute to the formation of O2. Production of
the production of O2, closing the oxygen mass balance (long NO is hardly observed over inert material (SiC, not shown).
dashed line in Fig. 6).
Further information on the relatively low amount of
Other Fe catalysts. Figure 8 shows that the promotion
NO2 formed compared to the amount of N2O decomposed effect of NO is not a unique feature of ex-framework Fe–
can be related to the decomposition reaction of NO2 to zeolite catalysts. Sublimed Fe/ZSM-5, Fe–ZSM-5, Fe-beta,
NO and O2. By feeding a mixture of NO2 in He over ex- and Fe–USY catalysts prepared by liquid (aqueous)-ion ex-
[Fe,Al]MFI(a), the NO2 decomposition rate was analyzed. change (Fig. 8a), as well as Fe supported on Al2O3 and SiO2
The result for p(NO2)o = 0.2 mbar is shown in Fig. 7. NO2 (Fig. 8b) all show a significant enhancement of the N2O con-
is converted into NO and O2 over ex-[Fe,Al]MFI(a) in the version when NO is added to the feed. The observed benefi-
temperature range where the promotion effect of NO oc- cial effect of NO is larger for Fe–zeolites than for Fe/Al2O3
curs. Equilibrium is reached above 700 K, which indicates and Fe/SiO2, manifested by a larger shift in operation
FIG. 6. NO2 (ꢂ) and O2 (ꢀ) formed vs N2O converted at 700 K over ex-[Fe,Al]MFI(a) at different partial NO pressures (mbar) in the feed.
Crosses (×) represent the NO2 + O2 measured experimentally and the diagonal dashed line the mass balance for oxygen. The numbers refer to the
NO/N2O ratio in the feed. Feed composition: 1.5 mbar of N2O + (0–1 mbar of NO) in He; P = 1 bar; GHSV = 60,000 h−1
.

´
´
216
PEREZ-RAMIREZ ET AL.
0.25
0.20
0.15
0.10
0.05
0.00
NO
x
175 K
p
NO2
650
550
750
T / K
850
950
FIG. 7. Partial pressure of (ꢁ) NO, (ꢀ) NO2, and (ꢃ) NOx vs tem-
FIG. 9. Partial pressure of (ꢁ, ꢂ) NO and (ꢀ, ꢃ) NO2 vs tempera-
perature over ex-[Fe,Al]MFI(a). Feed composition: 0.2 mbar of NO2 in
He; P = 1 bar; GHSV = 60,000 h−1. Equilibrium composition of NO and
NO2 (Eq. [4]) is represented by dashed lines.
ture over Fe-beta (circles) and Fe/Al2O3 (diamonds). Feed composition:
1.5 mbar of N2O + 0.2 mbar of NO in He; P = 1 bar; GHSV = 60,000 h−1
.
Equilibrium composition of NO and NO2 (Eq. [4]) is represented by
dashed lines.
temperature between the promoted and nonpromoted sys-
tems or by an increased N2O conversion at the same tem-
perature. The temperature shift in Fe–zeolites and Fe-
supported oxides for 50% N2O conversion is ∼100 and
∼50 K, respectively. In, e.g., Fe/ZSM-5 or Fe-beta, conver-
sion increases from 10 to 100% at 698 K in the presence
of NO, while the conversion over Fe/Al2O3 only increases
from 10 to 30% at 823 K. The ion-exchanged catalysts are
active in a temperature range similar to the ex-framework
catalysts (compare Figs. 1 and 8a). Promotion of N2O de-
composition and formation of NO2 over Fe/Al2O3 and
Fe/SiO2 take place at much higher temperatures. The
last observation is derived from the NOx profiles over
these catalysts, which are shown in Fig. 9 for Fe-beta and
Fe/Al2O3. Again the value of the partial NO and NO2 pres-
sure indicates that these are not controlled by the equilib-
rium in Eq. [4].
Multitrack Experiments
N2O decomposition. Multitrack was used to further in-
vestigate the formation of O2 during N2O decomposition
and the influence of NO in the N2O conversion (O2 forma-
tion) over ex-[Fe,Al]MFI(a). A typical Multitrack profile
during direct N2O decomposition at 823 K over this cata-
lyst after 100 pulses is presented in Fig. 10. The N2O and N2
signals are very similar in shape. The O2 response signal of
the N2O pulse, being much broader, deviates from the N2O
and N2 signal. In order to determine whether the shape of
(a)
(b)
1.0
1.0
0.8
0.6
0.8
0.6
X
X
0.4
0.4
0.2
0.2
0.0
0.0
550
650
750
850
600
700
800
900
T / K
T / K
FIG. 8. N2O conversion vs temperature over (a) low-temperature catalysts [(ꢆ, ꢇ) Fe/ZSM-5, (ꢄ, ꢅ) Fe–ZSM-5, (ꢀ, ꢁ) Fe-beta, (ꢃ, ꢂ) Fe–USY]
and (b) high-temperature catalysts [(ꢆ, ꢇ) Fe/Al2O3 and (ꢉ, ꢈ) Fe/SiO2]. Feed composition: open symbols, 1.5 mbar of N2O in He; solid symbols,
1.5 mbar of N2O + 0.2 mbar of NO in He; P = 1 bar; GHSV = 60,000 h−1
.

NO-ASSISTED N2O DECOMPOSITION OVER Fe CATALYSTS
217
NO-assisted N2O decomposition. To study the effect of
NO on the O2 formation in N2O decomposition, dual-pulse
experiments were carried out at different temperatures, in
which N2O and NO were pulsed at 0.1 and 1 s, respectively,
repeated in cycles of 2 s. The O2 response in different stages
of the experiment at 773 K is shown in Fig. 13. Pulsing N2O
(at 0.1 s) over ex-[Fe,Al]MFI(a) at this temperature leads
to a broad oxygen signal (Fig. 13, profile a), indicating a
relatively slow desorption of O2. Alternate pulsing of N2O
and NO for 10 cycles leads to a remarkable effect on the
O2 response at the time of the N2O pulse. The O2 response
sharpens (Fig. 13, profile b), indicating a significantly faster
O2 desorption in N2O–NO cycles than in the experiment
with only N2O. Stopping NO pulsing results in a transition
backtotheoriginalcatalystbehaviorinsixcycles(12s), with
a very broad O2 response (Fig. 13, profiles c and d). At lower
temperatures the promotion effect lasts longer. At 698 K,
the promotion is vanished 20 cycles (40 s) after stopping
NO pulsing. This indicates that adsorbed species formed at
the time of the NO pulse are involved in the enhanced O2
formation and their stability toward desorption determines
the duration of the effect. The sharp oxygen signal is rapidly
recovered after four pulses of NO (Fig. 13, profiles e and f).
During the dual-pulse experiments at these tempera-
tures, no NO or NO2 signals were observed at the time
of the N2O pulse, which indicates that NO does not block
N2O decomposition sites. NO2 formation was observed, but
only at the time of the NO pulse (Fig. 13, profile g). The MS
technique did not allow quantification of the NO2 desorp-
tion. At high temperatures (e.g., 973 K) the effect of NO
on the oxygen desorption behavior over ex-[Fe,Al]MFI(a)
is negligible (21), since the O2 profiles obtained by pulsing
N2O only, or N2O and NO sequentially, are very similar. It
should be noted that at such high temperature N2O conver-
sion over ex-[Fe,Al]MFI(a) was 100% both in the absence
and in the presence of NO under flow conditions.
FIG. 10. Typical Multitrack profiles of N2O, N2, and O2 during pulsing
of pure N2O (at 0.1 s) over ex-[Fe,Al]MFI(a) at 823 K. The upper profile
results from separate experiments pulsing pure O2 over the catalyst bed
at the same temperature. Profiles recorded after 100 pulses.
the oxygen is due to a mechanistic feature of N2O decompo-
sition or simply due to nonreactive adsorption–desorption
phenomena of O2 along the catalyst bed, molecular oxygen
was pulsed. In this case the pulse response is very similar
in shape to that of N2 and N2O (Fig. 10), confirming the
validity of the first hypothesis.
The delay of the oxygen signal was further investigated
by pulsing N2O at different reaction temperatures. These
results are shown in Fig. 11. In the range 573–673 K, a very
broad oxygen signal is observed. At 723 K the oxygen signal
becomes more pronounced. Around 873 K the signal sig-
nificantly sharpens, indicating that the rate of desorption of
oxygen from the catalyst surface is significantly enhanced.
DISCUSSION
Role of the Structure of the Fe Species in NO-Assisted
N2O Decomposition
The nature of the active sites in Fe–zeolites (especially
Valuable information on the dynamic oxygen desorption in FeMFI) prepared by various methods has already been
process from the catalyst surface is also obtained from the discussed extensively in the literature (24–29). Isolated iron
desorption profiles of O2 shown in Fig. 12. In these exper- ions or oligonuclear oxo-iron species in the zeolite channels,
iments, the O2 decay signal is continuously monitored af- as well as iron oxide particles (of different sizes and distri-
ter stopping N2O pulsing after 100 pulses at temperatures butions), are generally identified in these heterogeneous
between 673 and 798 K. At 673 K, surface oxygen slowly systems. The nature and relative amounts of the various
desorbs from the catalyst surface at a constant rate over species depend on the preparation method, which deter-
more than 25 s after stopping N2O pulsing. Desorption is mines the catalytic activity.
considerably faster at higher temperatures. At, e.g., 698 K,
The different activities of the catalysts in Figs. 1, 2, and
desorption takes ∼15 s, decreasing to ∼7 s at 723 K and so 8 should be related to the nature of the active site for N2O
on. At 798 K, oxygen desorption is relatively fast (∼5 s af- activation. The strength of the Fe–O bond formed in the
ter stopping N2O pulsing). This trend indicates the relative first step of the catalytic cycle (Eq. [1]) strongly depends
stability of the surface oxygen species with temperature.
on the nature of the iron species in the material, and here

´ ´
PEREZ-RAMIREZ ET AL.
218
FIG. 11. O2 profiles from Multitrack during continuous pulsing of pure N2O (at 0.1 s) over ex-[Fe,Al]MFI(a) at different temperatures. Profiles
recorded after 100 pulses.
the zeolite structure plays a crucial role in determining the
In view of the heterogeneous nature of the catalyst for-
specificity of these species. It is beyond the scope of the mulations with respect to iron, it is not surprising that the
present paper to discuss the exact nature of the active site promoting effect of NO on N2O decomposition has been
in N2O decomposition, although it is a highly challenging identified for zeolite catalysts of different constitutions
issue for catalysis.
(e.g., those prepared via an ex-framework route vs the
FIG. 12. Oxygen desorption profiles measured by Multitrack after stopping N2O pulses (N2O off) over ex-[Fe,Al]MFI(a) at different temperatures.
Pulses were stopped after 100 pulses of the gas, using a cycle time of 2 s.

NO-ASSISTED N2O DECOMPOSITION OVER Fe CATALYSTS
219
(g)
N2O on
NO on
(f)
(e)
(d)
(c)
(b)
(a)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
t / s
FIG. 13. O2 profiles measured by Multitrack during catalytic decomposition of N2O over ex-[Fe,Al]MFI(a) at 773 K using a cycle time of 2 s:
(a) pulsing N2O (at 0.1 s) after 100 cycles; (b) subsequent pulsing N2O (at 0.1 s) and NO (at 1.1 s) after 10 cycles; (c, d) pulsing N2O only (at 0.1 s),
three and five cycles after switching off the NO valve, respectively; (e, f) pulsing N2O (at 0.1 s) and NO (at 1.1 s), two and four cycles after switching
on the NO valve, respectively. (g) A typical NO2 profile during pulsing N2O (0.1 s) and NO (1.1 s) after 100 cycles.
catalysts prepared by ion-exchange or sublimation meth- has been attributed by Kapteijn et al. (20, 33) to scaveng-
ods). Even Fe catalysts on conventional supports (Al2O3, ing of adsorbed oxygen (O∗), regenerating the active site
SiO2), prepared by incipient wetness impregnation, also (Eq. [2]). In a recent paper we demonstrated that scav-
show an improved activity. In the latter catalysts iron is enging of adsorbed O∗ by NO is not the only mechanism
basically present as iron oxide particles (in a wide distribu- of promotion, and that NO also enhances O2 formation
tion of sizes). Apparently, the NO promotion is not specific (21), acting as a catalyst. Multitrack experiments presented
for a certain type of Fe species. This is most likely the result here show that oxygen desorbs from the catalyst surface in
of the high affinity of NO to reduced Fe sites, irrespective of a relatively slow step. Scheme 1 shows a typical reaction
the local structure of these sites. The high affinity of NO for pathway for catalytic N2O decomposition (12), including
reduced Fe sites has been demonstrated in various infrared the transition between the structures involved (Scheme 1,
studies, the absorption frequencies being assigned to dif- structures I–III). First, a vacant site (ꢆ) is oxidized by re-
ferent Fe configurations present in zeolites and located on action with N2O (reaction [1.1]), leading to Oꢆ species
conventional supports (21, 30–32). The order in activity of (unambiguously denoted as O∗). The reaction of a second
the different catalysts is not changed after addition of NO, N2O molecule with the oxidized site results in the regener-
which indicates a certain uniformity of the active sites pro- ation of the sites, according to reaction [1.2]. Oxygen des-
moted by NO. So, alternatively the activity and promotion orption can also take place by recombination of adsorbed
effect may be attributed to the same active species, present oxygen atoms (reaction [1.3]). In view of the slow oxygen
in different amounts in the various catalysts.
desorption (Figs. 11 and 12), the recombination of two oxy-
gen atoms (located on two different sites) appears the most
likely explanation for O2 formation. These oxygen atoms
can be originally separated, i.e., deposited at remote iron
Mechanism of NO-Assisted N2O Decomposition
N2O decomposition. As indicated in the Introduction, sites (Scheme 1, structure III). The dynamic nature of the
the beneficial effect of NO on the activity of FeZSM-5 adsorbed oxygen in the zeolite was studied by Hall and

´ ´
PEREZ-RAMIREZ ET AL.
220
NO pulses are needed to restore the promotion effect, the
pulse size, and the amount of catalyst used, and assuming
that all NO remains adsorbed at Fe sites, it can be calcu-
lated that less than 0.9% of the Fe present is involved in
the promotion. This also suggests that a very low fraction
of the Fe is active in the reaction.
Promotion by NO induces a significant decrease, ∼45–
55 kJ · mol−1, in the apparent activation energy for N2O de-
composition, which has also been observed in the reaction
of N2O and CO (with respect to direct N2O decomposition)
over ion-exchanged transition metal (Fe, Co, Cu) ZSM-5
catalysts (33), suggesting a similar effect on the reaction
mechanism, i.e., the removal of oxygen from the active site.
Whattherelativecontributionisofenhancedoxygendes-
orption vs the formation of NO2 over the catalysts needs
some further discussion. At relatively low temperatures the
first phenomenon is stronger than at high temperatures, as
indicated by the Multitrack analyses. Above 900 K the oxy-
gen response in the case of the ex-[Fe,Al]MFI(a) at the
N2O pulse is no longer affected by alternating N2O–NO
pulsing (21). To explain the enhancement of the O2 desorp-
tion rate at the time of the N2O pulse, an adsorbed species
formed at the time of the NO pulse needs to be involved.
The amount of adsorbed NO will be reduced at high tem-
peratures, while oxygen desorption at high temperatures
already proceeds fast. This adsorption involvement is sup-
ported by the slower decay of the NO promotion at lower
temperatures in the dual-pulse Multitrack experiments af-
ter stopping NO pulsing.
SCHEME 1. Schematic representation of a typical N2O decomposi-
tion mechanism.
coworkers (34) by step-switching experiments from N2O to
N218O over Fe–Mor. These studies showed that adsorbed
18O species (deposited by N218O) in Fe–Mor can exchange
with the lattice oxygen of the zeolite in the vicinity of the Fe;
so the oxygen must have a certain mobility to achieve this.
The slow oxygen desorption in Fe–zeolites can then be un-
derstood by a slow migration of adsorbed oxygen atoms.
The rate of desorption is higher at higher temperatures
(Figs. 11 and 12) and is attributed to a higher oxygen mobil-
ity and recombination. Once adsorbed atomic O∗ species
meet each other, desorption takes place rapidly, since no
O2 adsorption was observed for Fe–ZSM-5 (12, 35, 36).
The absence of the enhanced O2 desorption effect at high
temperatures is in agreement with the relatively smaller
promoting effect of NO on the performance of metal oxide
(Al2O3, SiO2)-supportedFecatalystsvszeoliticFecatalysts.
NO-assisted N2O decomposition. The rate of oxygen In the former catalytic systems, it is likely that removal of
desorption from the catalyst surface is greatly enhanced O∗ by NO, as stated in Eq. [2], plays a more important role
by NO addition. Indeed, a significant increase was found in the promotion than enhanced O2 desorption. Sang and
in the amount of oxygen formed if NO was added to the Lund (14) use the interconversion of nitrates and nitrites to
N2O-containing feed (Fig. 13), and only a relatively small explain the enhancement of N2O conversion by NO. How-
amount of NO is needed to dramatically increase the N2O ever, for Al2O3- and SiO2-supported Fe catalysts this expla-
decomposition rate (Fig. 3). Although during this process nation does not appear very likely, since nitrites and nitrates
NO2 is formed, even beyond the thermodynamic equilib- are not stable at the conditions where the promoting effect
rium between NO, O2, and NO2 according to Eq. [4] (see, is observed. Furthermore, in an infrared analysis of the Fe
e.g., Fig. 4), the enhanced oxygen production strongly sug- catalysts (21), nitrate bands have never been observed, only
gests a catalytic effect of NO. If Eq. [2] was the only pro- adsorbed NO and NO2 species.
motion route induced by the addition of NO, a progressive
Although NO2 is decomposed over Fe-based catalysts
increase in conversion upon increasing the inlet NO partial into NO and O2, the contribution of NO2 decomposition
pressure would be expected. This is not the case; the pro- to the formation of O2 in the NO-assisted N2O decompo-
motion occurs already at relatively low substoichiometric sition is likely to be limited. The Multitrack experiments
amounts of NO, indicating the catalytic nature, and reaches clearly show that oxygen desorption is triggered at the time
a limiting value at increasing molar NO/N2O feed ratios of the N2O pulse, and not at the time of the NO pulse.
up to 10. This suggests the involvement of NO adsorption, NO2 formation is observed at the time of the NO pulse,
and that the sites where NO is adsorbed are not in competi- indicating displacement of adsorbed NO2 by NO, as con-
tion with N2O decomposition sites. Competitive adsorption firmed by recent transient in situ FT-IR/MS studies over
would have resulted in inhibition, especially at high partial ex-[Fe,Al]MFI(a) (21). In the Multitrack, formation of O2
NO pressures. From the observation that at 698 K only four and NO2 is decoupled: NO2 is mainly released due to

NO-ASSISTED N2O DECOMPOSITION OVER Fe CATALYSTS
221
displacement by NO and probably by thermal desorption, Fe/SiO2. For low-temperature Fe–zeolite catalysts, N2O ac-
while O2 is formed during the N2O pulses. If NO2 decompo- tivation occurs next to an adsorbed NO molecule (∗NO).
sition, like that presented in Fig. 7, contributed significantly At these lower temperatures (550–700 K) the oxidized site
to O2 formation, an O2 response would have been expected (Oꢆ) subsequently oxidizes adsorbed NO to adsorbed NO2
at the time of the NO pulse, rather than at the time of the (reaction [2.3]), yielding structure III in Scheme 2. Subse-
N2O pulse only. This excludes Eq. [3] as a major cause of quently, NO2 desorption is induced by adsorption of NO
the promotion effect.
(reaction [2.4]), as was observed from the Multitrack ex-
periments by the presence of a NO2 signal at the time of
Reaction pathways and proposed mechanism. In view of the NO pulse. This closes the catalytic cycle for the conver-
the results presented in this paper and elsewhere (21), vari- sion of N2O and NO to N2 and NO2 at lower temperatures.
ous pathway(s) are needed to explain the catalytic effect of
Different options can be proposed to explain the en-
NO on the N2O decomposition. The different experimen- hanced oxygen formation. Again N2O activation occurs
tal observations are depicted in Scheme 2. The transitions next to an adsorbed NO molecule, yielding structure II in
of the different structures considered (Scheme 2, structures Scheme 2. Subsequently a second N2O molecule reacts with
I–IV) are also considered in the reaction mechanism. Ini- the site, yielding N2 and O2 (reaction [2.5]), and structure I
tially, N2O reacts on a vacant site, yielding N2 and leaving an is regenerated. The enhanced oxygen desorption from the
oxidized site (reaction [2.1]). On the catalyst, a substantial active center has been ascribed to a reduced stability of ad-
amount of adsorbed NO is present, as previously observed sorbed oxygen, induced by either electronic or steric effects
byinsituFT-IR(21). DuetotheabsenceofinhibitionbyNO of the NO adsorbed on neighboring oxidized sites (21). A
in the activity tests and the fact that no NO signal appears at more plausible explanation for the enhanced oxygen des-
the time of theN2Opulse inthe Multitrackexperiments, it is orption is reaction [2.6]. The increased N2O decomposition
concluded that the NO adsorption and N2O decomposition is simply explained by the recombination of oxygen present
do not compete for the same site, so both processes occur at in adsorbed NO2 and oxygen species deposited by N2O on
different Fe species (solid grey area and crosshatched area a neighboring site.
in Scheme 2). Adsorption of N2O (as O∗) and NO in two
Thus the adsorbed NO serves to accommodate tempora-
open coordinations at the same iron site seems hardly prob- rily the deposited oxygen from the N2O, freeing the
able, since dinitrosyls were not identified by infrared studies neighboring site for deposition of a second oxygen. This
in this system. The production of NO2 is shown in reaction would imply the presence of remote sites for the N2O
[2.2] (Scheme 2), which is likely the major mode of pro- decomposition. Consequently, the migration of oxygen
motion for high-temperature catalysts, i.e., Fe/Al2O3 and atoms to recombine to molecular oxygen constitutes the
SCHEME 2. Possible pathways in the NO-assisted N2O decomposition over ex-[Fe,Al]MFI(a).

´ ´
PEREZ-RAMIREZ ET AL.
222
The isolated nature of the sites prohibits the dissociation
of O2 explaining the absence of O2 inhibition. The forma-
tion of NO2 beyond thermodynamic equilibrium of Eq. [4]
indicates that Eq. [3] proceeds faster than Eq. [4] under
competitive conditions, where NO, N2O, O2, and NO2 are
simultaneously present. The NO2 formation is in fact due
to the displacement by NO, which adsorbs stronger (21).
Hence, O2 formation by adjacently adsorbed NO2 will be
suppressed under these circumstances, but could explain
the results of Fig. 7 in the absence of N2O. Alternatively,
O2 formation from NO2, via reaction [2.6], requires the
competition between N2O and NO2 to oxidize a vacant (ꢆ)
site. If N2O is a more efficient oxidizer of these sites, NO2
levels beyond thermodynamic equilibrium of Eq. [4] are
well feasible.
At this point it is worth noting two other catalyzed re-
actions where NO serves as an “atomic” oxygen transport
facilitator. The first is the old lead chamber process for SO2
oxidation, in which NO acts as a gaseous carrier (39). The
second is the role of NO in the bifunctional catalysis to oxi-
dize soot (40, 41). There a catalyst, e.g., metal oxide (Cr2O3,
Co3O4, CuO) or Pt based, oxidizes NO with O2 to NO2 and
the latter oxidizes the soot at temperatures where O2 is
not active. Although in these examples the oxygen is trans-
ported through the gas phase, in the current study this oc-
curs via an adsorbed mode. As mentioned before this prop-
erty of NO may be utilized in other (selective) oxidation
reactions.
SCHEME 3. Proposed NO-assisted oxygen desorption mechanism
during N2O decomposition.
rate-determining process in N2O decomposition. Taking
this further, it could be speculated that the presence of
NO on the catalyst surface enhances the mobility of ad-
sorbed oxygen via intermediate NO2 species. This mecha-
nism, schematically depicted in Scheme 3, leads to a faster
oxygen recombination, and thus an accelerated O2 desorp-
tion. The transfer of the oxygen atom from one NO to the
other is well possible, as previously observed for NO and
15N18O step changes over Cu–ZSM-5 and Cu–Y (34, 37).
Of course, the O2 formation by two adjacent adsorbed NO2
groups cannot be ruled out.
Finally the observed NO2 decomposition over ex-
[Fe,Al]MFI(a) and in general Fe–zeolites (reaction [2.7]
in Scheme 2) in flow experiments deserves some discus-
sion. It has been proposed that N2O decomposition over
Fe–zeolites yields the so-called α-oxygen (9), which has the
ability to selectively oxidize benzene in phenol. Since the
Fe–zeolites are able to convert NO2 into NO and O2 at
CONCLUSIONS
NO strongly promotes N2O decomposition over Fe-
relative high rates, i.e., equilibrium is attained at 700 K, it based catalysts. The promoting effect is observed for many
could be hypothesized on the basis of the proposed schemes different formulations and appears independent on the na-
that NO2 is also able to produce adsorbed (atomic) oxygen ture of the bulk of the Fe species in the catalyst. The tem-
species (from structure III to II in Scheme 2). So in princi- perature at which N2O decomposition occurs is strongly
ple it could be used as selective oxidant. A drawback, how- dependent on the catalyst. Small amounts of NO in the
ever, will be that the catalyst also effectively helps the O2 feed (NO/N2O < 0.25) are sufficient to produce a substan-
formation, thus rendering NO2 into a less efficient reactant tial increase in N2O conversion for all the catalysts. Higher
than N2O. Alternatively, mixtures of O2 and NO could be partial NO pressures (NO/N2O > 0.25) lead to a satura-
used.
tion behavior of the N2O conversion, although no inhibi-
The results in this paper have implications for the ap- tion by NO is observed, even at NO/N2O = 10. Apparently,
plication of N2O in selective oxidations. In the AlphOxTM different sites are involved in the NO adsorption and the
process (38), the N2O originates as a coproduct of the ox- deposition of oxygen by N2O. The latter sites seem to be
idation of cyclohexanone/cyclohexanol mixtures by nitric remote from each other, rendering the oxygen atom recom-
acid. This N2O contains also NO. It is clear that this NO has bination the rate-determining process. High temperatures
to be removed to large extent otherwise it will lower con- stimulate this process but also adsorbed NO at lower tem-
siderably the utilization of N2O in the benzene oxidation by peratures. Adsorbed NO may temporarily accommodate
providing an escape route for the so-called α-oxygen from oxygen deposited by N2O on a neighboring site, allowing
the N2O. This can be translated into an accelerated oxy- the deposition of a second oxygen and the recombination
gen desorption and/or the sequestration of reactive O∗ as of both. Less than 0.9% of the Fe seems to participate in
adsorbed ∗NOx species, impeding the effective transfer of this promotion, indicating the low density of active sites.
atomic oxygen to the benzene molecule.
Additionally, adsorbed NO may facilitate the migration of

NO-ASSISTED N2O DECOMPOSITION OVER Fe CATALYSTS
223
oxygen through NO2 intermediates, enhancing the change 20. Kapteijn, F., Mul, G., Marba´n, G., Rodr´ıguez-Mirasol, J., and Moulijn,
J. A., Stud. Surf. Sci. Catal. 101, 641 (1996).
21. Mul, G., Pe´rez-Ram´ırez, J., Kapteijn, F., and Moulijn, J. A., Catal. Lett.
of recombination to O2.
77, 7 (2001).
ACKNOWLEDGMENTS
22. Pe´rez-Ram´ırez, J., Berger, R. J., Mul, G., Kapteijn, F., and Moulijn,
J. A., Catal. Today 60, 93 (2000).
This research was financially supported by the Dutch Council for Chem-
23. Nijhuis, T. A., van den Broeke, L. J. P., Linders, M. J. G., Makkee, M.,
istry Research (CW-NWO). B. vd. Linden and S. Molla´-Romano are ac-
Kapteijn, F., and Moulijn, J. A., Catal. Today 53, 189 (1999).
knowledged for their assistance in the Multitrack measurements. Prof. Dr.
24. Lobree, L. J., Hwang, I.-C., Reimer, J. A., and Bell, A. T., J. Catal. 186,
Ir. A. Bliek is acknowledged for the use of the in situ FT-IR cell.
242 (1999).
25. Marturano, P., Drozdova´, L., Kogelbauer, A., and Prins, R., J. Catal.
REFERENCES
192, 236 (2000).
26. Battiston, A. A., Bitter, J. H., and Koningsberger, D. C., Catal. Lett.
66, 75 (2000).
27. Joyner, R., and Stockenhuber, M., J. Phys. Chem. B 103, 5963
(1999).
28. Long, R. Q., and Yang, R. T., J. Catal. 194, 80 (2000).
29. Pe´rez-Ramı´rez, J., Mul, G., Kapteijn, F., Moulijn, J. A., Overweg,
A. R., Dome´nech, A., Ribera, A., and Arends, I. W. C. E., J. Catal.
207, 113 (2002).
30. Lobree, L. J., Hwang, I.-C., Reimer, J. A., and Bell, A. T., Catal. Lett.
63, 233 (1999).
1. Feng, X., and Hall, W. K., Catal. Lett. 41, 45 (1996).
2. Chen, H.-Y., and Sachtler, W. M. H., Catal. Today 42, 73 (1998).
3. Ma, A.-Z., and Gru¨nert, W., Chem. Commun. 71 (1999).
4. Long, R. Q., and Yang, R. T., J. Catal. 188, 332 (1999).
5. Ko¨gel, M., Mo¨nnig, R., Schwieger, W., Tissler, A., and Turek, T.,
J. Catal. 182, 470 (1999).
6. Pophal, C., Yogo, T., Yamada, K., and Segawa, K., Appl. Catal. B 16,
177 (1998).
7. Centi, G., and Vazanna, F., Catal. Today 53, 683 (1999).
8. Mauvezin, M., Delahay, G., Kißlich, F., Coq, B., and Kieger, S., Catal.
Lett. 62, 41 (1999).
31. Hadjiivanov, K., Kno¨zinger, H., Tsyntsarski, B., and Dimitrov, L.,
Catal. Lett. 62, 35 (1999).
9. Panov, G. I., Uriarte, A. K., Rodkin, M. A., and Sobolev, V. I., Catal.
Today 41, 365 (1998).
32. Lezcano, M., Kovalchuk, V. I., and d’Itri, J. L., Kinet. Catal. 42, 104
(2001).
10. Dubkov, K. A., Sobolev, V. I., and Panov, G. I., Kinet. Catal. 39, 72 33. Kapteijn, F., Marba´n, G., Rodr´ıguez-Mirasol, J., and Moulijn, J. A.,
(1998).
J. Catal. 167, 256 (1997).
11. Ribera, A., Arends, I. W. C. E., de Vries, S., Pe´rez-Ram´ırez, J., and
Sheldon, R. A., J. Catal. 195, 287 (2000).
34. Valyon, J., Millman, W. S., and Hall, W. K., Catal. Lett. 24, 215
(1994).
12. Kapteijn, F., Rodr´ıguez-Mirasol, J., and Moulijn, J. A., Appl. Catal. B
9, 25 (1996).
13. El-Malki, E-M., van Santen, R. A., and Sachtler, W. M. H., J. Catal.
196, 212 (2000).
14. Sang, C., and Lund, C. R. F., Catal. Lett. 73, 73 (2001).
15. Long, R. Q., and Yang, R. T., Chem. Commun. 1651 (2000).
35. Sobolev, V. I., Panov, G. I., Kharitonov, A. S., Romannikov,
V. N., Volodin, A. M., and Ione, K. G., J. Catal. 139, 435
(1993).
36. Boreskov, G. K., in “Catalysis: Science and Technology” (J. R.
Anderson and M. Boudart, Eds.), Vol. 3, p. 39. Springer-Verlag, Berlin,
1982.
16. Pe´rez-Ram´ırez, J., Kapteijn, F., Mul, G., and Moulijn, J. A., Chem. 37. Valyon, J., and Hall, W. K., J. Catal. 143, 520 (1993).
Commun. 693 (2001). 38. Notte´, P. P. , Topics Catal. 13, 387 (2000).
17. Pe´rez-Ram´ırez, J., Kapteijn, F., Mul, G., and Moulijn, J. A., Appl. 39. Muller, T. L., “Kirk–Othmer Encyclopedia of Chemical Technology.”
Catal. B 35, 227 (2002).
(M. Howe-Grant, Ed.), 4th ed., Vol. 23, p. 363. Wiley, New York, 1997.
18. Oi, J., Obuchi, A., Bamwenda, G. R., Ogata, A., Yagita, H.,
Kushiyama, S., and Mizuno, K., Appl. Catal. B 12, 277 (1997).
19. Centi, G., Galli, A., Montanari, B., Perathoner, S., and Vaccari, A.,
Catal. Today 35, 113 (1997).
40. Mul, G., Zhu, W., Kapteijn, F., and Moulijn, J. A., Appl. Catal. B 17,
205 (1998).
41. Jelles, S. J., Makkee, M., and Moulijn, J. A., Topics Catal. 16/17, 269
(2001).